1. Field of the Invention
This invention relates to verifying data in a storage subsystem and more particularly relates to identifying and verifying the active data in the storage subsystem.
2. Description of the Related Art
A typical storage system, such as a redundant array of independent drives (“RAID”) structure, requires maintenance of the data in order to prevent data loss. One method for maintaining the data is to verify the data periodically. If data has been corrupted, the corrupted data can be isolated or recovered from redundant data. Data verification may be referred to as scrubbing.
The verification and correction process can become problematic in several ways. Often, the entire storage device is verified even if there is no useful data in many volumes, sectors or tracks of the storage device. Excessive storage subsystem resources are required to carry out the verification and correction process on all data in the storage system. Additionally, if the data written is not verified until the rest of the data on the storage device is verified, the exposure time of potentially corrupt data is unnecessarily prolonged.
To reduce the system requirements of verification and correction, processes have been developed that direct writes to a reduced portion of the storage system and then restrict verification to the reduced portion. However, the reduced portion strategy may reduce the storage system performance in other areas. In addition, data that is not written may also be corrupted. If data is modified or written to a storage device, data segments adjacent the sector being written may be corrupted. Data loss may also occur in a segment if corrupt data is rebuilt from redundant data that has also been corrupted.
Accordingly, a need exists for a process, apparatus, and system that verify only selected data in a storage system. Beneficially, such a process, apparatus, and system would reduce system resource requirements by reducing the quantity of data to be verified, while still increasing data reliability.